Shayan Pardis
Shayan Pardis
fixes #42 I undo the changes whenever inserting is not possible. However, the correct way to address this is to rebuild the filter with a different hash function which, I...
closes #473
closes #605
The functions to_cpu, to_gpu copy the data regardless of where it resides. It would be more intelligent to copy only if the data is not on the same device (calling...
In the current implementation, to_cpu, to_gpu copy the whole data. This means that `Adjoint` matrices get copied thus they are not adjoint anymore. As a result, the user has to...
In the current implementation, there is a lot of duplicate code in the GPU extension. The reason is that the `apply!` function uses indexing on rows of the Tableau thus...
## idea Assume that our GPU is capable of executing kernels of gates in parallel (running each on a different stream), or we have access to multiple GPUs and we...
Currently, we don't run GPU tests on CI/CD. In `test/.env` we have set `GPU_TESTS=false`. We should set that to `GPU_TESTS=true` on the server and have some GPU hardware to be...
On GPU there is an overhead whenever we call a new kernel. In `pftrajectories` function we are running a kernel for each gate. In the GPU kernel that we run...